Control rods are automatically scrammed by gravity when the power is cut. They are held by electro magnets on most designs. However, I admit that I'm not too familiar with GE's BWR.
When a U or Pu nucleus fissions there are several possible outcomes. Some of those outcomes yield radioisotopes whose half-life is on the order of a week. Thus when you suddenly interrupt a running nuclear reactor, you tend to still have a some energy output queued up to happen spontaneously regardless of any further fission/neutron chain reaction. IIRC, your "week" figure is roughly correct, but not because of ordinary heat capacity, but because of this continuing radioactive decay: the energy output typically continues for days to be several percent of steady state power output preceding the shutdown. (Which is enormously more energy than could be stored as heat in the mass of the reactor at any earthly temperature.)
And no, you CAN'T scram those because it's not a chain reaction going on but simply radioactive decay - there is no known force in the universe which can stop an unstable atom core from decaying.
I would image power would keep them out of the reactor and when power is cut, they drop in.
My ill informed understand of the Chernobyl accident is that specific safety measures had been deliberately disabled as part of testing.
Does anyone know why the Fukushima reactors weren't scrammed? Also, can you scram the waste pools?